Fix runaway event schedulers
It is not unusual to see instances of event-sched-new
on the ops node that have used hundreds of minutes of CPU. Sometime they will be actively using CPU and other times they will be quiescent. This is a real drag on the ops node.